The next stage of artificial intelligence can include human: It is a robot

propaganda

The next stage of Artificial Intelligence (AI) they RobotsWhich will help reduce the work world at work, as an executive official for NVIDIA announced the next “Euronws”. “We are in a very enjoyable moment. The promise of robots It is for a long time. It was in our imagination and science, “said” EURENEWS The Next “” Lebardian, Vice President of OMNIVERSE and Technology of NafidiaAt the Taiwan Computex Technology Fair.

Liparidian said that although technology companies are trying to build a public robot for general use for years, the problem was, despite their ability to build a material robot,, Programming was always a challenge. “I have changed all of this,” he says.

Companies like Tesla Rush to build human robots have made great progress. Last week, the company Elon Musk Declared his robot Optimus He learned to perform home tasks. However, robots still have a lot to learn.

In the case of NVIDIA, the company states that robots should learn their duties in the virtual world for security, but also because it will take a long time to train robots with humans. “The only way to create these smart robots is to use simulationLiparidian cream.

“he A basic problem What we have with IA material Thing IA is keen on data. You should feed the artificial intelligence factory with a lot and a lot of quality data to give you a vital training experience. “In the case of large linguistic models (LLM), there is a lot of data online to train them.

Robot training in data

However, in the case of physical artificial intelligence, many data cannot be extracted. “To get all the information we need to train a robot in a group of an object, we have to create it in one way or another,” he explains. “A collection of the real world is not possible. We cannot create enough data. Even if you can, in some cases it is dangerous, takes a long and expensive time. “

“A way to move from fossil data to renewable data sources,” Lebaridian says. The best renewable data for material data is Material simulationHe added.

Once the robot is tested, or “graduated” and it seems that it works well, it can go to the first employer. Recent graduation is trained with a group of Data available to the public. It is studied from textbooks and information that everyone can access everywhere. You have a general expert who enters your company, and they are useful.

“But it is not really useful until it is formed for some Years In your company’s information and data, which relate to your field and your own practices and how to do things, ”he added. In automatic phrases, this means that the user can The date was not specified in Human robots They will reach our lives, but they said it would be “soon.”

Where and what to use a robot

Los The first cases of use For them they will be in Factories. “I think it is Industrial use This will be the first, because even if we can build an ideal robot that you can use in your home, it is not clear that all human beings will love, “according to Libaridian.” But in the industry there is a great need for that. There are not enough young people to replace the qualified older workers who retire in all countries. “

the Global employment deficiency It has reached historically high levels in the past decade, according to the Organization for Economic Cooperation and Development. The descent of the population, as well as aging, and the fact that many people do not want to do “three d”, referring to the terminology in English, which, according to NVIDIA CEO, was “dangerous, boring and dirty works.”

Taiwan This mechanical need has joined and prepares to release a Quinquenal plan Last week, the government announced the strengthening of the automatic industry in an attempt to alleviate the lack of employment last week. The descent of Taiwan would test the country’s economy and its ability to serve the vulnerable and the elderly, according to Peter Hong, Director of the Ministry of Engineering and Technologies in the National Council for Science and Technology (NSTC), local media.

Liparidian said that after using it in factories, human robots can help in RetailSince many companies claim that they cannot employ caffeine individuals to accumulate shelves. He also said that it can be used in Minasand Nuclear reactors Or even in space. In the end, it can be used to care for older adults, if there is a request.

How to make the robot safe

But while we are excited about this new stage of artificial intelligence, Llm They are still wrong in many things, which sometimes makes them Things have been invented. Los The mistakes caused by a robot In the material world It can be more dangerous.

propaganda

However, Liparidian believes that like independent vehicles seem to start science fiction, people ultimately end up getting used to and improve technology. “In artificial intelligence, yes, there are still some inaccurate things, but I think we must admit that in the past two and a half years since its presentation Chatgptthe accuracy and quality What they also produce It increased significantly“, He said.

But he added that perhaps Chatbots will not be completely correct because we want humans to do tasks. “In reality There is no correct answer For many of these things, he said. “But for our tasks in industryThis is really something it is Very measurableFor example, did this object accurately collect and transfer it so far and do it safely and strongly? “

He also pointed out that these systems can Create and testChecking That is safe before implementing it. “We have machines and systems that we believe are very dangerous if they are not well configured. But we are able to create nuclear reactors and these systems, and keep them somehow. We can do the same with material artificial intelligence,” he says.